QOS is the performance management system within ONTAP.
It doesn’t just stop a volume from going above a certain level of performance,
it also monitors and manages performance. So it’s your one-stop shop for
most perf questions!
You can use ‘qos statistics
latency show’ to see
volume-level latency numbers.
-
The left most column ‘Latency’ is total latency
from request received until response to the client
-
‘Network’ is the time it takes to traverse
the network stack inside ONTAP
-
‘Cluster’ will show latency due to indirect IO
-
‘Data’ is ONTAP itself
-
‘Disk’ is time to
leave the controller and hit the disk
cluster1::> qos statistics latency show -iterations 100 -rows 3
Policy Group Latency Network Cluster Data Disk QoS
-------------------- ---------- ---------- ---------- ---------- ---------- ----------
-total- 110.35ms 110.02ms 0ms 327.00us 0ms 0ms
vs1vol0 167.82ms 167.22ms 0ms 603.00us 0ms 0ms
vol1 117.76ms 117.56ms 0ms 191.00us 0ms 0ms
vol2 44.24ms 44.05ms 0ms 190.00us 0ms 0ms
-total- 38.89ms 38.63ms 0ms 256.00us 0ms 0ms
vol2 64.47ms 64.20ms 0ms 266.00us 0ms 0ms
vol1 27.28ms 27.03ms 0ms 253.00us 0ms 0ms
vs1vol0 23.72ms 23.47ms 0ms 249.00us 0ms 0ms
-total- 409.81ms 409.65ms 0ms 169.00us 0ms 0ms
A simpler view with IOPS and throughput can be found using ‘qos statistics performance show.’
cluster1::> qos statistics performance show -iterations 100 -rows 4
Policy
Group
IOPS Throughput Latency
--------------------
-------- --------------- ----------
-total-
79 1296.00KB/s 337.41ms
_System-Best-Effort
25
0KB/s 0ms
vol1
24 96.00KB/s 193.72ms
vol2
18 1152.00KB/s 750.98ms
vs1vol0
12 48.00KB/s 707.38ms
-total-
109
1.99MB/s 133.27ms
Please note that the –total- line indicates the
system-wide performance. Very useful!
